Notice: Decentralized options figures are derived from a direct query to the DeFiLlama public API on August 12, 2026, with full daily series downloaded protocol by protocol. Deribit and Hyperliquid figures come from each platform's public APIs on the same day. Before using any recent data, we check for adapter lag, the results of which are in the third section. This is a dated snapshot: if you are reading this weeks later, refer back to the original sources. Nothing that follows constitutes financial advice or a recommendation for any platform. CleanSky does not receive commissions or referral payments from any of the platforms mentioned.
On July 1, 2026, Block Scholes and Castle Labs published the most comprehensive thesis on the renaissance of on-chain options, and their numbers are correct: the premium—what is actually paid to buy an option, not the notional value of the contracts—is at all-time highs. It is the conclusion that does not hold up. The renaissance exists, it is measurable, and it changes nothing, because it is occurring on a base so small that even tripling it leaves it within the margin of error of other markets. We audited the thesis using their same source and added two calculations of our own: that July 2026 saw 3.44x the premium of July 2025, and that the first twelve days of August 2026 are trending below the first twelve days of August 2025.
The Renaissance Thesis, and Why Its Numbers Are Correct
Let’s start by giving credit where it’s due. The Renaissance of Onchain Options relies on four observations, and all four are true.
The first is that the premium—the money actually paid to buy options, not the notional value of the contracts—is at all-time highs this year. Our download confirms this: March 2026 closed at $62.48 million in premium, the best month in the entire series. The second is year-over-year growth, and here we extend their series beyond the report's date: July 2025 moved $12.80 million in premium, while July 2026 moved $43.98 million. Multiplying by 3.44 in twelve months is our own calculation based on their source, confirming the direction described in the report.
The third is that infrastructure has genuinely improved. Rollups lowered transaction costs, order books replaced automated market makers—which never knew how to price volatility—and portfolio margin systems emerged. The fourth is that there are new entrants with real traction: Rysk on Hyperliquid, CallPut on Base, Paradex with perpetual options.
The report also acknowledges its own limits, which is what makes it serious: it explicitly states that liquidity remains fragmented across strike prices and maturities, and that on-chain platforms continue to be much smaller than traditional ones.
The discrepancy begins when measuring that 244% against the size of the market the sector aspires to disrupt.
How Much Do On-Chain Options Move Compared to Deribit and Hyperliquid?
These are the three figures for August 12, 2026, all representing 24-hour notional value and all from primary sources.
| Market | 24h Notional ($ millions) | Source |
|---|---|---|
| On-chain options, all 28 protocols combined | 27.36 | DeFiLlama |
| Deribit, BTC and ETH options only | 1,344.03 | Deribit API |
| Hyperliquid, perpetuals | 3,044.12 | Hyperliquid API |
Decentralized options represent 2.04% of Deribit and 0.90% of a single perpetuals exchange. To put it in perspective without a calculator: one day of Hyperliquid is equivalent to one hundred and eleven days of the entire on-chain options category combined. Or, put another way, what the entire sector moves in a day, Hyperliquid moves in about thirteen minutes.
Daily data is fragile, and it must be noted: on-chain notional fluctuated between $7.07 and $61.54 million in the preceding seven days. That is why we repeated the calculation over thirty days, where the daily average rises to $46.25 million and the median sits at $31.49 million. Compared to the same day on Deribit, the full range goes from 2.0% to 3.4%. None of the three readings change the order of magnitude, and it is the order of magnitude that sustains the argument.
There is a second methodological warning regarding the on-chain figure. The aggregate returned by the API, $26.43 million, mixes the last available data from each adapter between August 11 and 12; the $27.36 million in the table is the protocol-by-protocol sum for August 12 itself. The difference is 3.5% and does not alter any conclusions, but both figures are in circulation and it is worth knowing which is which.
And Deribit is no ordinary rival. Its open interest in options that same day was $24.949 billion between BTC and ETH. In other words, the open interest of a single centralized platform is about nine hundred times the daily volume of all decentralized ones combined. These scales are not even remotely comparable.
The difference between premium and notional matters here, because both are cited as "volume" but do not measure the same thing. Notional is the value of the underlying asset covered by the contracts; premium is what actually changes hands. In the on-chain category, premium was 2.03% of notional over the last thirty days—a numerical coincidence with the previous ratio, which measures something else. When someone says on-chain options move $1.4 billion a month, they are citing notional: the real money was $28.22 million.
Is It Still Growing? The First Twelve Days of August, 2026 vs. 2025
This check contradicts the 244% extrapolation, which is why it precedes the discussion on mechanisms.
If the year-over-year growth is 3.44x, one would expect August 2026 to crush August 2025. It isn't: it's trending lower. And for the comparison to be honest, one must compare the same calendar days, not twelve days against an entire month.
| Period | Total Premium ($ millions) | Daily Median ($ thousands) |
|---|---|---|
| August 1 to 12, 2025 | 8.13 | 807 |
| August 1 to 12, 2026 | 5.93 | 383 |
| June 2026 (30 days) | 55.51 | 1,333 |
| July 2026 (31 days) | 43.98 | 1,042 |
We use the median rather than the mean intentionally. July had days of $3.1 and $3.8 million that pull any average upward, and a month with two peaks and twenty-nine sluggish days is poorly represented by the mean. This is the same measurement problem we documented in the on-chain credit ranking, where three days of liquidations distorted an entire ranking.
The daily median for the first twelve days of August 2026 is $383,000. For the same twelve days in 2025, it was $807,000: the start of this August is at 47.5% of where it was twelve months ago, following a year of triple-digit growth.
Before assigning value to that, we checked the obvious: that it wasn't an artifact of adapters that hadn't closed the day. Derive, Aevo, Panoptic, and Hypersurface had their last data on August 11; Rysk V12, CallPut, Paradex, and Hegic were up to date on the 12th. A one-day lag at most, over a twelve-day series. The drop is real.
What we won't do is turn twelve days into a trend. It could be August holidays, it could be a low-volatility regime reducing hedging demand, or it could be the end of the rally. The honest approach is to state what observation would resolve this: if the daily median for September 2026 returns above $1 million, August was seasonality; if it stays below $700,000, the March peak was the cycle top and not the beginning of anything.
How Many Options Protocols Are Still Alive? Twelve Out of Twenty-Eight
Of the 28 options protocols listed by DeFiLlama, these are the ones still operating.
DeFiLlama lists 28 options protocols. There are twelve with volume in the last thirty days. There are eight with volume in the last twenty-four hours. The remaining sixteen show zero: Premia V2 and V3, Stryke CLAMM, Smilee, Moby, Valorem, Arrow Markets, IVX, Jasper Vault, PancakeSwap Options, Racks, Toros, Umoja Synths, Kyan Blue, and the old versions of Rysk and Derive.
And of the twelve that are breathing, one takes almost everything.
| Protocol | 30-Day Notional ($ millions) | Share |
|---|---|---|
| Derive | 1,111.77 | 80.1% |
| Rysk V12 | 105.83 | 7.6% |
| CallPut | 90.41 | 6.5% |
| Paradex | 45.58 | 3.3% |
| Aevo | 21.10 | 1.5% |
| Panoptic V2 | 6.92 | 0.5% |
| Remainder (22 protocols) | 5.96 | 0.4% |
A sector where a single protocol concentrates 80.1% of the volume is not a sector; it is a company with decorative competition. If Derive were to have a problem tomorrow, the entire category would disappear from the dashboards.
There is a detail that summarizes the matter better than any aggregate figure. Panoptic is the only genuinely original design in the group: perpetual options built on Uniswap liquidity positions, with no expiration and no upfront premium: the strike price is set by the liquidity range, not a listed contract. In other words, the innovation consisted of resembling a perpetual. It moves an average of $230,800 a day, 0.5% of a category that is already just 0.90% of Hyperliquid. Architectural originality did not buy liquidity.
Why Do Options Disperse Liquidity While Perpetuals Concentrate It?
The first reason for the structural ceiling is geometric, not technological.
A Bitcoin perpetual is one order book. Everyone who wants leveraged exposure to BTC ends up in the same place, so all depth is stacked at a single point. If you need to understand the internal mechanics, you can find them in how perpetual funding rates work.
A Bitcoin option is not an instrument: it is a surface. Every combination of strike price and expiration date is a different contract with its own book. On August 12, 2026, Deribit had 794 BTC instruments live at once, and another 670 for ETH. The same amount of money spread across 794 books is 794 times thinner in each one.
This cannot be fixed with incentives or more users: it stems from the product's form. Perpetuals concentrate liquidity by design and options disperse it by design, which is why the perpetual won in crypto even before decentralized options had a product to show. If you are interested in the general mechanism, we elaborate on it in the slippage simulator: the depth of a book determines what it costs you to exit, and a thin book charges you for the exit even if the screen price looks good.
Why Can't a Market Maker Quote On-Chain?
The second reason for the ceiling is operational, and it explains why the platforms that survived stopped being what they claimed to be: Derive, Aevo, and Paradex—the three that moved the order book off-chain—account for 84.9% of the category's notional as of August 12, 2026.
Whoever prices an option doesn't just quote and leave. They must continuously re-quote, because the contract's value changes with the underlying price, implied volatility, and the simple passage of time. Furthermore, they must hedge: every time the market moves, they adjust their directional exposure by buying or selling the underlying. It is a continuous operation. If the concept sounds familiar, it is the same one we explain in what it means to be delta neutral.
On a blockchain, every re-quote is a transaction: it costs money, takes a block, and is public before it is executed. A market maker who re-quotes a thousand times a day cannot pay for a thousand transactions a day, nor can they afford to have their intent read before it is confirmed.
The solution the sector found was to move the order book off-chain and leave only the settlement on-chain. This is what Derive, Aevo, and Paradex do. It works—in fact, it is the reason they are the ones left alive—but we must say out loud what it implies: order matching, which is where price discovery happens, occurs on private servers. What is decentralized is custody and settlement, not the market. Those who choose these platforms out of distrust for intermediaries end up with their orders matched on the servers of Derive, Aevo, or Paradex.
Why Does Portfolio Margin Decide Where Professional Capital Operates?
The third reason is capital efficiency, and it is the one that carries the most weight for those moving large sizes: the $24.949 billion in open interest sustained by Deribit on August 12 exists because a unified account backs them with a fraction of the collateral that the on-chain version would require.
If you sell a fully collateralized covered call, you lock up the entire underlying asset. If you set up a spread—selling one option and buying another to limit loss—the real risk you run is the difference between the two, but a system that collateralizes each leg separately requires collateral for both.
Portfolio margin calculates collateral based on the net risk of all your positions at once. And the next step, which is where the real advantage lies, consists of crossing that margin with your perpetual and spot books in the same account: hedging one position reduces the collateral requirement for the other. That is a unified account, and it is exactly what a centralized exchange can offer and an on-chain platform finds very difficult to replicate without reintroducing a custodian.
The consequence is direct. Those who sell volatility professionally need to rotate capital, and between a place that locks up their notional and another that locks up their net risk, there is no decision to make. The best customer of the options market is structurally incompatible with the on-chain version of the product.
What Happened to Option Vaults Like Ribbon Finance?
There is an important precedent, as it was the only time decentralized options had real money.
Between 2021 and 2022, option vaults appeared, automating the systematic sale of volatility: the depositor would put in their ETH and the vault would sell covered calls every week. Ribbon Finance reached over $300 million in deposits. The problem was design: the vaults auctioned at the same time, on the same day of the week, and with predictable strike prices, so the market learned to anticipate them and systematically bought cheap volatility from them. In 2023, Ribbon merged into Aevo and pivoted to a derivatives exchange with perpetuals. The flagship project of on-chain options ended up becoming a site for trading perpetuals.
And the institutional demand that was theoretically supposed to reach crypto options did arrive, but through a different door. The renaissance report itself notes that options on BlackRock's ETF surpassed BTC options on Deribit in open interest. This is an uncomfortable fact for the thesis citing it: it means that when institutional capital wanted Bitcoin options, it bought them in a regulated traditional finance wrapper.
Corporate movement points in the same direction. Coinbase agreed in May 2025 to buy Deribit for approximately $2.9 billion—$700 million in cash and 11 million Class A shares—and closed the deal that same August. At the time of the agreement, Deribit handled around 85% of the global crypto options market; in the first half of 2026, that share is around 49%, split mostly with Bybit, Binance, and OKX. Options liquidity did not just stay centralized: it consolidated even further, and now belongs to a U.S. listed company.
What Can't We Prove and What Would Make Us Change Our Minds?
This analysis has two limits, which we declare upfront.
The first is source coverage. The entire count of protocols—the twelve alive and the sixteen at zero—depends on DeFiLlama having an adapter for every platform that exists. If there are options platforms without an adapter, they are excluded from our count. And a protocol showing zero could be dead or could have a broken adapter; from the outside, these two situations look the same, and we have not been able to distinguish them case by case.
The second is volume quality. We cannot determine what part of Derive's volume is organic and what part responds to incentive programs, as that breakdown is not public. If a significant fraction were incentivized, the picture would be even worse than what we describe. We state this in that direction intentionally: our conclusion does not depend on resolving this doubt.
A thesis that does not state how it can be refuted is an opinion. Ours is that the ceiling for on-chain options is structural rather than one of adoption. Three things would overturn it:
- Concentration decreases. If in six months Derive has less than 50% of the notional and there are at least four protocols above 10%, the sector will have stopped being one company with sidekicks.
- Real on-chain cross-margin appears, without a custodian, between options and perpetuals. This is the mechanism that today only a centralized exchange can offer, and whoever solves it without reintroducing trust will have eliminated the most difficult advantage to copy.
- Monthly premium moves from tens of millions to hundreds. As long as the figure remains in the range of July's $44 million, any growth percentage is being calculated on a base that doesn't move the needle on anything.
In the meantime, the answer to whether it is worth analyzing this market as a market is: not yet. As a case study on why liquidity concentrates where it does, however, it is one of the best in crypto: here you can count exactly how many pieces the money is split into: 794 BTC books on a single platform and 28 protocols of which only eight are breathing.
Related articles: Decentralized Perpetuals in 2026. How Perpetual DEX Funding Rates Work. Hyperliquid vs. GMX vs. dYdX. What It Means to Be Delta Neutral. Crypto Options Platform Comparison. Track your portfolio on CleanSky — no yield promises, just your data.